Efficient production of deproteinization rubber using orthogonal experiments coupled with artificial neural networks

Deproteinized natural rubber variants were synthesized through modifications in enzymatic hydrolysis parameters and centrifugation settings. A neural network model was trained based on orthogonal experimental outcomes to analyze the impact of enzymatic hydrolysis temperature, duration, pH, enzyme quantity, centrifugation time, and speed on the tensile strength of deproteinization rubber. Predictions for the tensile strength of deproteinization rubber were made within an acceptable margin of error. The experimental findings demonstrate the efficacy of utilizing orthogonal experiments in conjunction with neural networks to forecast results and minimize the need for extensive experimentation.